The purpose of this research is to facilitate the detection, mapping and ultimate identification of genes that cause diseases in man. Statistical methods of family and pedigree analysis will be further developed for this purpose and implemented in computer programs. Methods of segregation, linkage and association analysis will be extended in various directions, as determined by the needs of genetic epidemiologists. Two particular types of models for the genetic analysis of pedigree data, regressive models and the finite polygenic-mixed model, have complementary strengths and weaknesses. They will therefore both be developed and investigated, with special attention being given to disease traits, possibly with variable age of onset. Segregation, linkage and association ("candidate gene") analysis will all be considered together, reflecting the current needs of genetic epidemiology and current capabilities of computer technology. Theoretical developments will continue to include studies of validity, power and robustness of all procedures developed, and special attention will be given to how studies can be most efficiently designed to unravel the genetic determinants of multifactorial diseases. The most efficient designs will probably necessitate sampling families in special ways, and hence require the development of new methodology to extract the maximum amount of information from these specially ascertained families. There will be ongoing collaborations, comprising the analysis of family and pedigree data collected by others, to ensure that the theoretical developments are relevant to current problems in the genetic analysis of multifactorial traits.